Samenvatting

This paper investigates the role of conversational AI in improving the quality of decision-making across different stages of the process. Drawing on an experimental study involving 113 students in Dutch higher education, participants engaged in a simulated business challenge using three approaches: no AI, delegative AI, and conversational AI. The study applied Horngren’s five-step decision-making model (Datar & Rajan, 2021) to evaluate perceived outcome quality.
Findings show a consistent progression in decision quality: from no AI, to delegative AI, and further improvement with conversational AI. The most significant gains were reported in the early stages, problem definition and solution generation, where iterative interaction and critical reflection are essential. Conversational AI was seen to enhance assessment by encouraging dialogue, refining ideas, and increasing clarity. While delegative AI improved task efficiency, conversational AI contributed more substantially to the depth and coherence of decision outcomes.
These insights are highly relevant for public governance. As institutions adopt digital technologies to improve citizen engagement, conversational AI may support more inclusive and reflective policy processes (Wirtz et al., 2019). Rather than automating decisions, it facilitates deliberation by simulating human-like dialogue, enabling users to explore alternatives and clarify underlying values (Rahwan et al., 2019). Previous research suggests that dialogic systems can promote trust, improve transparency, and mitigate algorithmic opacity (Shin & Park, 2021; Binns et al., 2018).
This study contributes to debates on AI in public and organizational contexts by demonstrating how interactive AI can support policy processes and meaningful citizen engagement in governance. While the research was situated in education, its implications extend to participatory governance and improve decision-making, particularly in settings where dialogue, contextual sensitivity, and shared understanding are essential to the legitimacy and quality of decision-making.